5.3. Redefinition of some national hardness reference scales As a result of international hardness comparisons, or on other reasons, in some cases a redefinition (shift) of a national hardness reference scale may become necessary. Three known cases deserve mention here. a) HRC hardness testers produced in the Soviet Union in the 1930's were checked by using imported blocks calibrated by the Wilson Company in the USA [8-8]. HRC values measured in the Soviet industry corresponded accordingly to the Wilson Reference scale of that time. The use of this HRC scale in industry was maintained until 1980. After the second World War Soviet Metrolaogical Institutes ect up deadweight Rockwell. standard machines which were compared with standards of other states.In this way the values of the so-called All-Union Standard were conform with other national standards, but differed from hardness values measured in the Soviet Industry. This anomaly was removed in 1979 by publishing the State standard Specification GOS? 8064-79 establishing the legal status of the All-Union Rockwell standard machine and of the hierarchy scheme of Rockwell hardness measurements performed in the Soviet Union. The Standard contains a conversion table between the tradilioual industrial Kockwell values (denoted by WRC) and the Rockwell values of the standard equipment (denoted by URC,). Some characteristic points of the conversion tabli ERC, HRC 20.0 17.8 30.0 28.1 40.0 38.4 50:0 48.7 60.0 59.0 65.0 64.1 GOST 8064 specifies also some practical measures to ensure the transition to hardness values conform to international practice [K- 14): Hardness testers used in industry are to be modified; hardness values specified in drawings or other technical documents mist be given in HRC. values; blocks calibrated in HRC values can be used with conversion tables during the two years of validity of their calibrated value. b) In the United Kingdom, in order to @isseminate to industries a standard HRC scale with a level that is commercially viable whilst retaining the stability of the national reference scale, a small corrections was applied to the hardness values assigned to each hardness block calibrated on the NPL deadweight machine [W-23]. The corrections, of less than 1 ARC unit, have been evaluated from a knowledge of the effect of design features in commercial machines which tend to reduce the hardness values, together with an allowance for the slightly harder values obtained when using shorter dwell times of only a few seconds compared with a standardizing time of 30 seconds. The effect of these technical developments is that HRC values with 70traceability to NPL in 1982 were about 1 HRC unit lower than they were in the 1970's. This is in agreement with practice in other major industrial countries. ©) The effect of dwell time and of other test conditions on hardness reference scales was examined also by the ‘echnical Committee IS0/1C 164. The changes introduced in the new series of 180 International Standards on hardness testing, elaborated in the 1980's, may have an effect on national hardness reference scales. Earlier loading cycles employed in industrial hardness testing and at calibrating hardness test blocks were different. In industry the hardness test should be perfurmed as quickly as possible, of course without excessive loss of precision. For block calibration the metrology laboratory has more time, consequently load rise time and dwell times were specified so that the penetration process be completely stabilized, to ensure maximum precision and accuracy During the elaboration of the revised International Standards on hardness testing the practical opinion prevailed and was fixed as specification for the loading cycle: The calibration of the blocks should be performed in a loading cycle which is identical with that employed in hardness tests in industry. What are the possible effects of this decision on existing hardness reference scales [P-26]? For the calibration of Rockwell blocks we are specially concerned with the following elements of the loading cycle: ~ to, time of application of the preliminary test force ty, load rise time from preliminary test force to total test force - tyr time of application of the total test force. The values given in the old and new ISO specifications for the calibration of standardized blocks for Rockwell hardness testing machines are shown in Table 10. Table 10 Values specified in the Standards 180/R 674 180 674 (1968) (2988) te 10 to 20 s 1 to 10 s # 2lo 6s te 30 to 35 s 3to 5s Time of application of the preliminary test force (t.) ‘The former practice was to maintain the preliminary force for 10 to 20 s, because approximately 10 s was necessary to stabilize the position nof the indenter. The new prescription is 1 to 10 s. DAMBACHER found a difference of 0.3 to 0.4 HRC in the final hardness value of a block of 20 HRC, depending on whether the preliminary test force was maintained for 1 s or 10 s. This tolerance range for time. in the unstabilized part of the indentation cycle is very wide for calibration purposes. Load rise time (t,) The difference of hardness values obtained at the two ends of the range specified for load rise time (2 and 8 s, respectively) is 0.2 to 0.3 HRC at 65 HRC, and 0.2 HRC at 20 HRC. ‘Dime of application of the total test force (t, ‘The value specified earlier was 30 te 35 s, and has now become 3 to 5s. Shorter application times mean higher hardness values. This effect is greater on materials of low hardness. Stabilization of penetration is not to be expected before 30 s. Table 11 shows hardness values obtained with different application times. Table 11 Hardness values (HRC) obtained with a duration of 30 s 5s 3s 20.0 20.6 to 20.7 20.7 to 20.9 60.0 60.2 60.3 This means that the national hardness reference scales should be corrected by up to 0.9 HRC at the low end and by 0.3 HRC at the high end of the HRC-scale. Even the limiting values of the newly specified range (3 to 5 s) may result in a hardness difference of 0.1 to 0.2 BRC at 20 HRC 0.1 HRC at 60 HRC Table 12 shows a similar effect in the case of the HRB scale. Table 12 Hardness values (HRB) obtained with a duration of 156 5s a8 55.0 56.1 56.8 Addition of uncertainties resulting from relatively large range of time specifications ‘Three kinds of uncertainties, originating from the specified ranges of time have been mentioned so far (Table 13). 72The resulting uncertainty can be considered in two ways: A) Tet us assume, firstly, that timo valuec within the specified ranges will be employed by calibration personnel at random. In this case a standard deviation can be estimated from the sum by quadratic addition, by using the appropriate formula. the three specified time ranges result in an increase in the uncertainty of the hardness standardizing machine by 0.4 HRC. b) We may assume, alternatively, that a given national standardizing institute will use only a narrow part of the ranges specified in the Slandard. But these will not necessarily be indentical to those used by other national institutes. The result will be a systematic difference between the calibrated values obtained by different institutes which may be as great as the arithmetic sum indicated in Table 13, namely 0.7 HRC. The international uniformity of hardness measurements is thus impaired. Table 13 Range of hardness values (in HRC) resulting from the range at hardness level of time specified for 20 ERC 60 HRC te 0.3 to 0.4 0.2 ta 0.2 0.2 to 0.3 tm 0.1 ta 0.9 ee Sum by simple addition 0.7 Sum by quadratic addition 0.4 6. Uncertainty of industrial hardness testers The final aim of all metrological work connected with hardness testing is the metrological assurance of industrial hardness testers. A detailed discussion of problems of these hardness testers goes beyond the scope of the present work. Only some questions are mentioned here, with hints and references for those who intend to study this field more in detail. Several decades ago, the first step towards the unification of hardness measurements was the survey of the state of hardness testers used in various laboratories, lw eslablishement of their precision and relative measurement deviation (accuracy). Extensive comparisons were made, the essential conclusions of which form now the basis of ISO International Standards on the verification of hardness testers (180 716- 1986, ISO 1079-1989, ISO 146-1984, ISO 156-1982 for the Rockwell, Vickers and Brinell methods). These Standards can be regarded as the condensed outcome of many years of experience. ‘The metrological behaviour of hardness testers is specified in the 13standards by two characteristics: Repeatability, i.e. the range of five consecutive measurements made on a Standardized block. Error i.e. the difference between the mean of the five measurements and the specified hardness of the standardized block. Fig. 28 illustrates this specification on the example of Rockwell testers at the levels of 60 and 20 HRC respectively. The range specified for the error is identical (+ 1.5 HRC) at all hardness levels. The range for repeatability depends on hardness. H is the hardness of the slandardized block used. Hi, H2,..+Hs are the tive measured hardness values, Hi the mean of the five. Specifications given in a standard represent minimum requirements necessary for the use of the hardness tester. In special application more severe requirements, or other specifications better suited for the given task may be employed. In such cases some of the oxperience of earlier research work may be useful, such as HILD [H-4] on the examination of Rockwell testers, JEGADEN et al. [J-2] and YAMASHIRO et al. [¥-10] On the examination of Super-Rockwell testers, PILIPTCHUK [P-14] on the theoretical foundations of the fine adjustment of hardness testers, ROSSOW [R-I3] on the examination of Rockwell ¢ testers; with well founded desian of exparimants and aualnatian of reenlte Surveys on the state of industrial hardness testing in various countries can be found in the publications enumerated in Table 14. Table 14 State of hardness testing in given countries Author Reference country Year of publication| Meyer yeaa FR. Germany 1957 Rossow R-12 FR. Germany 1961 Yoshizawa ya Japan 1961 Rossow R13 F.R. Germany 1967 Catka 3 Czechoslovakia 1970 Yano et al ¥-26 Japan 1972 Slinozuka-matsuura] — S-18. Japan 1973 Eckardt-Kersten E-1 GOR 1974 Shinozuka s-19 Japan 1975 Darbacher D-1 F.R. Germany 1978 The real problems facing industrial hardness testing can be better understood by the following considerations, by the help of an imaginary example on the heat treatment of a machine component [8-2] 14Buyer’s tolerance range Manufacturer’ s solerance range Mean = 58 HRC specified value at the manufacturer 2 anes Nominal ‘value eee) Nominal value | Satactures* by manufacturer™| [toy buyer reed tolerance range fas_ reduced toler— qs] Fig. 29. Frequency distribution of the hardness ~~ of heat treated components. 18On Fig. 29 the frequency distribution of the hardness of a batch of heat treated components is shown, which is supposed to be Gaussian. The standard deviation of measured hardness values was found to be + 1 HRC. The value specified on the drawing of the component is (58 + 2)HRC. Let us suppose that the mean of measured hardness values corresponds to the prescribed value, namely 58 HRC (what is of course an optimistic assumption). In this case the tolerance range of + 2 HRC includes 95 % of all produced pieces, 5 % (area shaded by points) is rejected by the manufacturer. The pieces are delivered to another company whose hardness tester has a systematic deviation of 0.5 HRC with respect to that of the manufacturer. This means the rejection of another 5 % of the delivered Pieces (shaded area uuder Lhe curve) when control tests are made. So far the possible errors of hardness testers were not fully taken into consideration. According to ISO 716 the error of the hardness tester may be + 1.5 HRC with respect to the block of the verification officer. ‘Thus in our example, in an extreme case, the systematic difference between measurements made by the manufacturer and the receiver may amount to 3 ARC. This may result in disputes which can be resolved only by direct hardness comparisons between the two partners. Another problem is the random error of the testers. ISO 716 specifies @ repeatibility of 1.2 HRC at 58 HRC, i.e. in the case of five repetitions of the hardness measurement, the difference between the maximum and minimum value may be 1.2 HRC. This corresponds to a standard deviation of + 0.5 HRC. If the manufacturer is cautions, desiring to avoid disputes, he reduces the product tolerance of + 2 HRC to 4 1.5 HRC 50 as to take into account the uncertainty of the tester Th thic cace (Fig. 29 b) already 14 % of heat treated pieces are to be rejected, in Place of the 5 % mentioned earlier. 16REFERENCES (The numbering employed in the Bibliography "The metrology of hardness scales" - 1981 was retained in the present publication). B-2 Barbato G., Desogus S. : Realizzazione di alcune scale di durezza Presso 1'Istituto di Metrologia "G. Colonnetti" di Torino. Ingegneria meccanica 1979, No. 4, 61-84. B-5 Bochmann G., Hild K.: Der Einfluss von Abweichungen des Eindringkdrpers von der Sollform auf dic Hirtewerte bei dei Messverfahren nach Rockwell C. Zeitschrift fir Instrumentenkunde 68 (1960), 155-162. B-10 Barbato G.: Investigations on the uncertainty of hardness measurement. Bulletin OIML No. 91 - Juin 1983, 3-12. B-14 Barbato G., Desogus S., Li F., Weng Q.: Comparison of hardness measurements between China (NIM) and Italy (IMGC). Bulletin OIML No. 105, 1986, 18 23. B-15 Barabanow S.P., Kotchin 0.M.: Gosudarstvenniy spetsialniy etalon edinits tverdosti po shkalam Brinellya. Izmeritelnaya Tehnika, 1987, No. 5, 3-4. B-16 Baker G.M., Yamamoto H., Yamamoto T.: Investigation of the indicated values of hardness blocks made in the USA and in Japan. Journal of material Testing Research Association of Japan, 33/1988/, No. 2, 111- 216 (avaliable in knglash). c-1 Gutka B.: Messeigenschaften der Rockwell-Hirtenormalgerdite Typ wazau (Halle). Feingerdtetechnik 15 (1966), 569-575. (More in detail, in Czech language: Mérova Technika 7 (1968), 17-26). €-3 Gutka B.: Die Genauigkeit der Hirtepriifung in der tschechoslowakischen Industrie. Feingeriitetechnik 19 (1970), 404-405. ©-6 Gutka B.: Metrology of Vickers hardness (in Czech language). Eeskoslovenskd standardizace, 1964, 253-255. C7 Gotha B.: Metrology of Rockwell. - A,B,C hardness (in Caech language) Coskoslovenskd standardizacc. 1905, 159-168, c-9 Gutka B.: the Czechoslovak State Standard of the Rockwell-N,T hardnesa scales (in Czech language). Ceskoslovenska ctandardizace, 1900, 187- i3i. C-11 Gutka B.: Possibilities of the international unification of hardness scales (in Czech language). Accurate Measurement and Metrology of Hardness, CSVTS, Praha, 1988, 96-116. Cr r—“—“‘_‘eesmsmmhMhUMmMm—™— (in Czech language). Mérovd technika 1965, 96-103. 7Danbacher H.: Fragen ur Normung der Hartepriifverfahren und Hartepriifgeriite nach Brinell, Rockwell und Vickers. VDI-Berichte Nr. 308, 1978, 77-90. DeBellis A.: Maintaining Rockwell Hardness Standards. Materials Research and Standards, Yol. 12, No. 8, 1972, 28-30. Eckardt H., Kersten W.: Stand der Makroeindruck-Hirtemessung in der DDR-Industrie. Standardisierung und Qualitat 20 (1974). 344-347. Eckardt H., Kersten W.: Interaction between indenters and HRC standard machines. Personal communication on experiments performed in ASMW, Berlin in 1982. Flurschiitz F.: Einige Betrachtungen tber Normal-Hlrtepriifgerate und die Aufstellung einer Hurtebezugsbasis. Feingeratetechnik 15 (1966), 564-569. Frank K.: Die Bedeutung der Kontrollplittchen bei der Hartepriifung. Werkstattstechnik 31 (1937), 213. Gidel A.Yu.: Method opredeleniya pogreshnostey formi nakonechnikov dlya tverdomerov. Ismeritelnaya technika 1967, No. 3, 33-38. Gosset A.: Intercomparaison des échelles de dureté HRC et HV 30 entre la France et la Honarie. Bulletin RNW. Na 49, mctchre 1090, 23-43. Hild K.: Der Einfluss der Toleranzen des Messverfahrens nach Rockwell C auf die Hartewerte. Zeitschrift flr Instrumentenkunde 66 (1958), 202-207, 230-234. Hild K.: Untersuchungen ber den Einfluss der Gestalt von Eindringkérper und Probe in der Hértemessung nach Rockwell-c und Vickers, VDI-Berichte Nr. 43, 1961, 47-50. Hormith K.: Der Einfluss der Belastungszeit bei Rockwell- Hartemessungen hoher Genauigkeit. VDI-Berichte Nr. 41, 1961, 105-108. Hormith K.: Das Normalgerlt des Deutschen Amtes ftir Messwesen zur Bestimmung der Vickershérte. Feingerdtetechnik 11 (1962), 250-253. Hormath K.: Mesurages internationaux de comparaison de dureté et proposition pour la fixation d'une base de référence pour le procédé de mesurage de dureté Rockwell. Bulletin de 1'OIML No. 14 (1963), 21- 30. Hormith K.: Zum gegenwirtigen internationalen Stand auf den Gebieten Ger Kraft- und Hartemessung. Feingerdtetechnik 15 (1966), 482-491. Helmrath Ch., Gould D.: The development of a unified European HRC hardness scale (also in German). VDI-Bericht Nr. 804, 309-374 18I-1 Tiguka K., Imai H.: Form measurement of small spherical surfaces by means of the method of least squares. Bulletin of the Japan Soc. of Prec. Engg-, Vol. 4, No 1, 1-6 I-2 Tiguka K., Yano H., Hida N., Goto M., Yamamoto K.: An international comparison of hardness conversion between Vickers and Rockwell C scales. Proceedings of the Round Table Discussion on Hardness Testing. ‘7th INEKO, London, 1976, 89-105. I-5 Tiguka K.: A comment on the uncertainty of hardness measurement. 9th IMEKO Congress, TC 5 Round Table, Berlin-West, 1982. J-2. Jegaden G., Siener P., Voinchet J.: Die Priifung der Harte von Blechen. vpI-Berichte Nr.100, 1967, 77-87. K-14 Kotchin 0.M., Koslow V.I.: 0 vvedenii etalonnoy shkali tverdosti Rokvella. Ismeritelnaya tehnika, 1983, No. 3, 48-49. K-18 Kotchin 0.M.: Sovremennoye metrologitcheskoye obespetcheniye ismereniy tverdosti. In: Sovremenniye metrologitcheskiye problemi fisiko- technitcheskih ismereniy. Moskwa 1988. K-21 Kovées G.: The effect of correction values at international hardness comparisons. Mérésiigyi Kozlemények XXVII (1986), 88-94. M-1 Marriner R.S.: Performance tests on indentars for Rockwall hardness testing. Engineering, July 1, 1955, 17-19. M-4 Marriner R.S., Wood J.G.: Developments in the accurate measurement of indentation hardness. Joint Convention on machines for materials and environmental testing, 6-10th Sept. 1965, paper 10, 79-84, Inst. mech. Engrs. N-5 Marriner R.S.: Leistungsteste an Eindringkirpern, die der Rockwell- Harteprifung dienen. Microtecnic XV (1961), 279-288. M-9 Marriner R.S., Wood J.G.: Comparison of international Rockwell C and Vickers HV 30 hardness scales during 1970 and 1971. Bulletin of OIML, No 48, Sept. 1972, 19-31. M-11 Meyer K.: Das Erreichen der Vergleichbarkeit der Hértemesswerte - ein problem der Werkstoff-Priiftechnik von internationaler Bedeutung. VDI- Berichte Bd.11, 1957, 103-122. M-24 Meyer K.: Die Eingliederung des Hirtepriifverfahrens nach Rockwell-C in das betriebliche Mess- und Priifwesen. Werkstattstechnik, Nov. 1952. N-26 Mikoszewski Twardosciomerze (Hardness testers. In Polish), Warszawa 1985. 19Pz Po3 Poa P12 OIML SP 19-Sr 4: Compte rendu de comparaison internationale des échelles de dureté. Prague 1984. Petik F.: Diskussionsbeitrag. 1. Internationaler Erfahrungsaustausch Kraft- und Hartemessung, Berlin, 1966, 18-20, Petik F.: Uber die Genauigkeit des ungarischen _Normal- Hartepriifgerdtes nach Rockwell. VDI-Berichte Nr. 100, 1967, 157-159. Petik F., Kovacs G.: Experiments to ensure the reproducibility and conformity of national hardness scales. ACTA IMEKO 1967, Vol. I, 369- 373. Petik F., Cutka B., Hadjiski I., Kersten W., Mikoszewski J. Vergleichsmessungen zwecks Vereinheitlichung der Hartebezugsskalen von fUnf Léndern. Vortrige vom V. Kongress fiir Materialpriifung, Budapest 1970, Bd. T, 163-168. (A more detailed version in Hungarian in: 137-143). Finommechanika 11/1972/, Petik F,: Uber einen Vorschlag flr eine einheitliche Methode zur Festlegung der Grésse der Unsicherneit beim Ubertragen der Harte— Besugsskalen mittels Hirtevergleichsplatten. VDI-Berichte Nr. 160, 1972, 205-203. Petik F., Marriner R.S.: Systematic differences of two standard scales of Rockwell C hardness as established by repeated comparisons. Bulletin of OIML No. 49, Dec. 1972, 21-25. Petik F. Wechselwirkung zwischen Hlrtenormalger’t und Normaleindringkérpern bei Hartemessungen nach Rockwell Cc. Materialpriifung 15 (1973), 239-241. Petik F.: On the choice of the confidence level for statistical tests of systematic differences. Proceedings of the Symposium on Measurement Theory, Enschede, Dec. 1975, 1-11. Potik F.: Reliability of the determination of systematic differences between measurements by different Rockwell C indenters. Proceedings of the Round Table Discussion on Hardness Testing. 7th IMEKO, London 1976, 83-88. Petik F.: Zur Metrologie der Harteskalen. Feingeréitetechnik 26 (1977), 175-177, Petik F.: Die staatliche Sicherung der Einheitlichkeit der Hartemessung. VDI-Berichte Nr. 308, 1978, 71-76. Philipps C.E., Fenner A.J.: Analysis of the Rockwell hardness test, employing a new deadweight testing machine. Engineering, July 13, 1951, 57-60. Piliptchuk B.I.: Pogreshnosti ismereniya tverdosti po Rockwellu. ‘Trudi WNIIN 50 (110), Moskwa-Leningrad 1961, 5-21. 80